Bert-base-multilingual-cased-finetuned-amharic Open-source Model - More Accurate Amharic Named Entity Recognition

Bert Base Multilingual Cased Finetuned Amharic

Developed by Davlan

Amharic language model fine-tuned based on multilingual BERT, achieved by vocabulary replacement and fine-tuning, outperforming the original multilingual model in NER tasks

Large Language Model

Transformers

#Amharic BERT #Multilingual fine-tuning #Named Entity Recognition

Downloads 196

Release Time : 3/2/2022

Model Overview

This model is a BERT model fine-tuned on Amharic corpus, specifically optimized for Amharic text processing tasks, with excellent performance in Named Entity Recognition tasks

Model Features

Amharic-specific vocabulary

Replaced the original mBERT's unsupported Amharic vocabulary, significantly improving the understanding of Amharic text

NER task performance advantage

Achieved an F1 score of 60.89 on the MasakhaNER dataset, significantly outperforming the original multilingual BERT model

Efficient fine-tuning

Efficient fine-tuning based on the pre-trained multilingual BERT model, saving training resources

Model Capabilities

Amharic text understanding

Masked word prediction

Named Entity Recognition

Use Cases

Natural Language Processing

Amharic text analysis

Processing Amharic news articles and other text content

Achieved an F1 score of 60.89 in NER tasks

Language model applications

Used for tasks such as masked word prediction in Amharic

🚀 bert-base-multilingual-cased-finetuned-amharic

This is an Amharic BERT model. It replaces the mBERT vocabulary with Amharic vocabulary and fine - tunes the bert - base - multilingual - cased model on Amharic texts, offering better performance in named entity recognition.

🚀 Quick Start

You can use this model with Transformers pipeline for masked token prediction.

>>> from transformers import pipeline
>>> unmasker = pipeline('fill - mask', model='Davlan/bert - base - multilingual - cased - finetuned - amharic')
>>> unmasker("የአሜሪካ የአፍሪካ ቀንድ ልዩ መልዕክተኛ ጄፈሪ ፌልትማን በአራት አገራት የሚያደጉትን [MASK] መጀመራቸውን የአሜሪካ የውጪ ጉዳይ ሚንስቴር አስታወቀ።")

✨ Features

bert - base - multilingual - cased - finetuned - amharic is an Amharic BERT model. It replaces the mBERT vocabulary with Amharic vocabulary due to the lack of language support and fine - tunes the bert - base - multilingual - cased model on Amharic language texts. It provides better performance than the multilingual Amharic on named entity recognition datasets. Specifically, this model is a bert - base - multilingual - cased model fine - tuned on the Amharic corpus using Amharic vocabulary.

💻 Usage Examples

Basic Usage

>>> from transformers import pipeline
>>> unmasker = pipeline('fill - mask', model='Davlan/bert - base - multilingual - cased - finetuned - amharic')
>>> unmasker("የአሜሪካ የአፍሪካ ቀንድ ልዩ መልዕክተኛ ጄፈሪ ፌልትማን በአራት አገራት የሚያደጉትን [MASK] መጀመራቸውን የአሜሪካ የውጪ ጉዳይ ሚንስቴር አስታወቀ።")

📚 Documentation

Intended uses & limitations

How to use

You can use this model with Transformers pipeline for masked token prediction. The code example is shown above in the "Basic Usage" section.

Limitations and bias

This model is limited by its training dataset of entity - annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.

Training data

This model was fine - tuned on [Amharic CC - 100](http://data.statmt.org/cc - 100/)

Training procedure

This model was trained on a single NVIDIA V100 GPU

Eval results on Test set (F - score, average over 5 runs)

Dataset	mBERT F1	am_bert F1
[MasakhaNER](https://github.com/masakhane - io/masakhane - ner)	0.0	60.89

BibTeX entry and citation info

By David Adelani

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご